1. Technical Field
The present disclosure relates to the field of sensor networks and the alerts they develop when sensing emergent information that they are cued to look for.
2. Description of the Related Art
Currently, system sensors collect data in a non-intelligent manner. That is, even if a sensor has limited intelligence (e.g., a camera that automatically tracks moving objects), most of the data collected by the sensors, and then transmitted to a controller, is meaningless. That is, sensors typically transmit data in a continuous manner, such that most of the transmitted data is “dead air” in which nothing of interest is happening. Even if some intelligence is present in the sensor, the data transmitted concerns what that particular sensor type is able to detect, and thus has little selectivity associated with such transmitted data. Thus, most sensors systems are either burdened with high positive and negative error rates, or send “dead air,” or both. To find a subject matter of interest, the controller must perform extensive data mining, using programs that search for patterns of interest in the massive amounts of previously stored data. Most searching is for simple, single sensor type, threshold events.